Direct air capture of CO2 for solar fuel production in flow

Abstract

Direct air capture is an emerging technology to decrease atmospheric CO₂ levels, but it is currently costly and the long-term consequences of CO₂ storage are uncertain. An alternative approach is to utilize atmospheric CO₂ on-site to produce value-added renewable fuels, but current CO₂ utilization technologies predominantly require a concentrated CO₂ feed or high temperature. Here we report a gas-phase dual-bed direct air carbon capture and utilization flow reactor that produces syngas (CO + H₂) through on-site utilization of air-captured CO₂ using light without requiring high temperature or pressure. The reactor consists of a bed of solid silica-amine adsorbent to capture aerobic CO₂ and produce CO₂-free air; concentrated light is used to release the captured CO₂ and convert it to syngas over a bed of a silica/alumina-titania-cobalt bis(terpyridine) molecular-semiconductor photocatalyst. We use the oxidation of depolymerized poly(ethylene terephthalate) plastics as the counter-reaction. We envision this technology to operate in a diurnal fashion where CO₂ is captured during night-time and converted to syngas under concentrated sunlight during the day.

Keywords: Carbon capture and storage; Photocatalysis; Solar fuels.

Fig. 1. DACCU through a dual-bed flow reactor consisting of DAC and CO₂U units.

a, Schematics of the system during light-off night operation. b, Schematics of the overall system during light-on day operation. c, The carbon capture unit with chemical CO₂ capture and release equations. d, The solar-driven CO₂U unit material composition and the relevant reduction and oxidation reactions. RT, room temperature; MFC, mass flow controller; PEI, polyethyleneimine; PET, poly(ethylene terephthalate); EG, ethylene glycol.

Fig. 2. DAC and solar-driven photothermal release of CO₂.

a, CO₂ levels in the outflow during DAC and the adsorbed CO₂ amount over time. DAC was performed with 600 mg of SBA-15|PEI adsorbent with an airflow rate of 90 ml min^–1 at room temperature. b, Photothermal CO₂ desorption setup with photothermal coating and parabolic trough reflector (note that the light source is for demonstration purposes only and different from the actual solar simulator used). c, CO₂ concentration in outflow gas stream during release with different flow rates under 3 suns of solar irradiation with photothermal cover. In all cases, the adsorbent bed temperatures reached around 85–100 °C across different regions under concentrated light during desorption. For a, data are presented as the average of two independent runs and the individual data points are shown in hollow circles. Source data

Fig. 3. Characterization and performance of the CO₂U unit.

a, HAADF-STEM image and the relevant EDS maps (O, Si, Ti, N) of the composite. b, The effect of different alcohol electron donors on H₂ and CO formation. Reactions were carried out in batch with 100 mg of CO₂U composite (nSiO₂|TiO₂|CotpyP) and 0.15 ml of electron donor under 1 sun (100 mW cm^–2, AM 1.5G) for 20 h. c, The effect of N₂ versus CO₂ carrier gas on product formation in fixed-bed flow setup. Reactions were performed with 250 mg of CO₂U composite, moistened with 0.50 ml EG, under 1-sun illumination at a carrier gas flow rate of 1 ml min⁻¹. d, The effect of different support materials on performance under similar conditions. e, An image of the parabolic trough reflector used for light concentration with the mounted tubular fixed-bed reactor. f, The effect of increasing solar intensity (around 3 suns) via reflector on CO formation when using γ-Al₂O₃|TiO₂|CotpyP as the CO₂U composite. g, The effect of temperature regulation on CO and H₂ formation. The 43 ^oC and 56 ^oC reactions were performed without temperature regulation (elevated temperatures due to solar thermal heating), whereas in the 25 ^oC reaction, temperature was kept steady using a water jacket. h, Product evolution with time under 3-sun illumination when temperature was kept steady at 25 °C. For b–d and f–h, data are presented as the average of two independent runs and the individual data points are shown in hollow circles. Source data

Fig. 4. CO₂U oxidation product analysis and dilute CO₂ response.

a, The overall reduction and oxidation products observed (γ-Al₂O₃|TiO₂|CotpyP as CO₂U composite and EG as electron donor, reaction time 12 h). b, Total charge carriers involved in product formation. c, Long-term product formation activities and CO selectivity when PET-derived EG was used as reductant (for the EG derivation protocol from PET, see Methods). Reaction was performed at a CO₂ flow rate of 3 ml min⁻¹. d, The effect of CO₂ dilution on CO generation in different carrier gases (CO₂ in nitrogen, air). All reactions were performed under 3-sun illumination at 25 ^oC temperature. For a, b and d, data are presented as the average of two independent runs and the individual data points are shown in hollow circles. Source data

Fig. 5. DACCU to produce solar syngas.

a, The designed reactor for in-flow DAC and CO₂U. b, CO₂ levels and photocatalytic CO and H₂ formation rates in the outflow during DACCU. The CO₂ values during capture (0–12 h) are multiplied by a factor of 200 for ease of visualization. DAC was performed during light-off operation at room temperature under airflow (90 ml min⁻¹) and the CO₂U is carried out during light-on operation under 3-sun illumination at 25 ^oC under N₂ flow (1 ml min⁻¹). c, Cumulative H₂ and CO yield over time during DACCU. d, Recapture and rerouting of unreacted CO₂ for increased conversion and low carbon emission. The additional downstream unit was loaded with fresh adsorbents for unreacted CO₂ capture and rerouting during conversion. e, Cumulative H₂ and CO formation over time during first- and second-pass conversion during operation under simulated solar irradiation. Data are presented as the average of two independent runs, and the individual data points are shown in hollow circles. Source data